src/language/control/define.c

   1 /* PSPP - a program for statistical analysis.
   2    Copyright (C) 2021 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation, either version 3 of the License, or
   7    (at your option) any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program.  If not, see <http://www.gnu.org/licenses/>. */
  16
  17 #include <config.h>
  18
  19 #include <limits.h>
  20
  21 #include "language/command.h"
  22 #include "language/lexer/lexer.h"
  23 #include "language/lexer/macro.h"
  24 #include "language/lexer/scan.h"
  25 #include "language/lexer/token.h"
  26 #include "libpspp/message.h"
  27
  28 #include "gl/xalloc.h"
  29
  30 #include "gettext.h"
  31 #define _(msgid) gettext (msgid)
  32
  33 static bool
  34 force_macro_id (struct lexer *lexer)
  35 {
  36   return lex_token (lexer) == T_MACRO_ID || lex_force_id (lexer);
  37 }
  38
  39 static bool
  40 match_macro_id (struct lexer *lexer, const char *keyword)
  41 {
  42   if (keyword[0] != '!')
  43     return lex_match_id (lexer, keyword);
  44   else if (lex_token (lexer) == T_MACRO_ID
  45            && lex_id_match_n (ss_cstr (keyword), lex_tokss (lexer), 4))
  46     {
  47       lex_get (lexer);
  48       return true;
  49     }
  50   else
  51     return false;
  52 }
  53
  54 /* Obtains a quoted string from LEXER and then tokenizes the quoted string's
  55    content to obtain a single TOKEN.  Returns true if successful, false
  56    otherwise.  The caller takes ownership of TOKEN on success, otherwise TOKEN
  57    is indeterminate. */
  58 static bool
  59 parse_quoted_token (struct lexer *lexer, struct token *token)
  60 {
  61   if (!lex_force_string (lexer))
  62     return false;
  63
  64   struct substring s = lex_tokss (lexer);
  65   struct string_lexer slex;
  66   string_lexer_init (&slex, s.string, s.length, SEG_MODE_INTERACTIVE, true);
  67   struct token another_token = { .type = T_STOP };
  68   if (string_lexer_next (&slex, token) != SLR_TOKEN
  69       || string_lexer_next (&slex, &another_token) != SLR_END)
  70     {
  71       token_uninit (token);
  72       token_uninit (&another_token);
  73       lex_error (lexer, _("String must contain exactly one token."));
  74       return false;
  75     }
  76   lex_get (lexer);
  77   return true;
  78 }
  79
  80 int
  81 cmd_define (struct lexer *lexer, struct dataset *ds UNUSED)
  82 {
  83   if (!force_macro_id (lexer))
  84     return CMD_FAILURE;
  85
  86   /* Parse macro name. */
  87   struct macro *m = xmalloc (sizeof *m);
  88   *m = (struct macro) {
  89     .name = ss_xstrdup (lex_tokss (lexer)),
  90     .location = xmalloc (sizeof *m->location),
  91   };
  92   *m->location = (struct msg_location) {
  93     .file_name = xstrdup_if_nonnull (lex_get_file_name (lexer)),
  94     .first_line = lex_get_first_line_number (lexer, 0),
  95   };
  96   lex_get (lexer);
  97
  98   if (!lex_force_match (lexer, T_LPAREN))
  99     goto error;
 100
 101   size_t allocated_params = 0;
 102   while (!lex_match (lexer, T_RPAREN))
 103     {
 104       if (m->n_params >= allocated_params)
 105         m->params = x2nrealloc (m->params, &allocated_params,
 106                                 sizeof *m->params);
 107
 108       size_t param_index = m->n_params++;
 109       struct macro_param *p = &m->params[param_index];
 110       *p = (struct macro_param) { .expand_arg = true };
 111
 112       /* Parse parameter name. */
 113       if (match_macro_id (lexer, "!POSITIONAL"))
 114         {
 115           if (param_index > 0 && !m->params[param_index - 1].positional)
 116             {
 117               lex_error (lexer, _("Positional parameters must precede "
 118                                   "keyword parameters."));
 119               goto error;
 120             }
 121
 122           p->positional = true;
 123           p->name = xasprintf ("!%zu", param_index + 1);
 124         }
 125       else
 126         {
 127           if (lex_token (lexer) == T_MACRO_ID)
 128             {
 129               lex_error (lexer, _("Keyword macro parameter must be named in "
 130                                   "definition without \"!\" prefix."));
 131               goto error;
 132             }
 133           if (!lex_force_id (lexer))
 134             goto error;
 135
 136           if (is_macro_keyword (lex_tokss (lexer)))
 137             {
 138               lex_error (lexer, _("Cannot use macro keyword \"%s\" "
 139                                   "as an argument name."),
 140                          lex_tokcstr (lexer));
 141               goto error;
 142             }
 143
 144           p->positional = false;
 145           p->name = xasprintf ("!%s", lex_tokcstr (lexer));
 146           lex_get (lexer);
 147
 148           if (!lex_force_match (lexer, T_EQUALS))
 149             goto error;
 150         }
 151
 152       /* Parse default value. */
 153       if (match_macro_id (lexer, "!DEFAULT"))
 154         {
 155           if (!lex_force_match (lexer, T_LPAREN))
 156             goto error;
 157
 158           /* XXX Should this handle balanced inner parentheses? */
 159           while (!lex_match (lexer, T_RPAREN))
 160             {
 161               if (lex_token (lexer) == T_ENDCMD)
 162                 {
 163                   lex_error_expecting (lexer, ")");
 164                   goto error;
 165                 }
 166               char *syntax = lex_next_representation (lexer, 0, 0);
 167               const struct macro_token mt = {
 168                 .token = *lex_next (lexer, 0),
 169                 .syntax = ss_cstr (syntax),
 170               };
 171               macro_tokens_add (&p->def, &mt);
 172               free (syntax);
 173
 174               lex_get (lexer);
 175             }
 176         }
 177
 178       if (match_macro_id (lexer, "!NOEXPAND"))
 179         p->expand_arg = false;
 180
 181       if (match_macro_id (lexer, "!TOKENS"))
 182         {
 183           if (!lex_force_match (lexer, T_LPAREN)
 184               || !lex_force_int_range (lexer, "!TOKENS", 1, INT_MAX))
 185             goto error;
 186           p->arg_type = ARG_N_TOKENS;
 187           p->n_tokens = lex_integer (lexer);
 188           lex_get (lexer);
 189           if (!lex_force_match (lexer, T_RPAREN))
 190             goto error;
 191         }
 192       else if (match_macro_id (lexer, "!CHAREND"))
 193         {
 194           p->arg_type = ARG_CHAREND;
 195           p->charend = (struct token) { .type = T_STOP };
 196
 197           if (!lex_force_match (lexer, T_LPAREN)
 198               || !parse_quoted_token (lexer, &p->charend)
 199               || !lex_force_match (lexer, T_RPAREN))
 200             goto error;
 201         }
 202       else if (match_macro_id (lexer, "!ENCLOSE"))
 203         {
 204           p->arg_type = ARG_ENCLOSE;
 205           p->enclose[0] = p->enclose[1] = (struct token) { .type = T_STOP };
 206
 207           if (!lex_force_match (lexer, T_LPAREN)
 208               || !parse_quoted_token (lexer, &p->enclose[0])
 209               || !lex_force_match (lexer, T_COMMA)
 210               || !parse_quoted_token (lexer, &p->enclose[1])
 211               || !lex_force_match (lexer, T_RPAREN))
 212             goto error;
 213         }
 214       else if (match_macro_id (lexer, "!CMDEND"))
 215         p->arg_type = ARG_CMDEND;
 216       else
 217         {
 218           lex_error_expecting (lexer, "!TOKENS", "!CHAREND",
 219                                "!ENCLOSE", "!CMDEND");
 220           goto error;
 221         }
 222
 223       if (lex_token (lexer) != T_RPAREN && !lex_force_match (lexer, T_SLASH))
 224         goto error;
 225     }
 226
 227   struct string body = DS_EMPTY_INITIALIZER;
 228   while (!match_macro_id (lexer, "!ENDDEFINE"))
 229     {
 230       if (lex_token (lexer) != T_STRING)
 231         {
 232           lex_error (lexer, _("Expecting macro body or !ENDDEFINE"));
 233           ds_destroy (&body);
 234           goto error;
 235         }
 236
 237       ds_put_substring (&body, lex_tokss (lexer));
 238       ds_put_byte (&body, '\n');
 239       lex_get (lexer);
 240     }
 241   m->location->last_line = lex_get_last_line_number (lexer, 0);
 242
 243   macro_tokens_from_string (&m->body, body.ss, lex_get_syntax_mode (lexer));
 244   ds_destroy (&body);
 245
 246   lex_define_macro (lexer, m);
 247
 248   return CMD_SUCCESS;
 249
 250 error:
 251   macro_destroy (m);
 252   return CMD_FAILURE;
 253 }
 254
 255 int
 256 cmd_debug_expand (struct lexer *lexer, struct dataset *ds UNUSED)
 257 {
 258   settings_set_mprint (true);
 259
 260   while (lex_token (lexer) != T_STOP)
 261     {
 262       if (!lex_next_is_from_macro (lexer, 0) && lex_token (lexer) != T_ENDCMD)
 263         {
 264           char *rep = lex_next_representation (lexer, 0, 0);
 265           msg (MN, "unexpanded token \"%s\"", rep);
 266           free (rep);
 267         }
 268       lex_get (lexer);
 269     }
 270   return CMD_SUCCESS;
 271 }